Thunderbolt Driver For Apple MBP Systems Gets Updated

[phoronix]

Phoronix: Thunderbolt Driver For Apple MBP Systems Gets Updated

Experimental Linux kernel code for providing a Thunderbolt driver that works on Apple MacBook Pro systems has been updated and is still hopeful for its eventual mainline kernel inclusion...

http://www.phoronix.com/vr.php?view=MTY2MDg

[wizard69]

This is good to hear even though I generally run Linux on non Apple hardware.

In any event I wonder why other systems would rely upon firmware as has been suggested in this article.

[smitty3268]

In any event I wonder why other systems would rely upon firmware as has been suggested in this article.

That's the way Intel wrote the spec. Apple did their own thing, as usual.

[chithanh]

Apple controls both Firmware and the Operating System on Macs, so they can put the hot-plugging code where it causes the least pain for them. As long as it "just works" in OS X, almost all of their users won't mind.

[litfan]

That's the way Intel wrote the spec. Apple did their own thing, as usual.

What's "as usual" supposed to mean exactly? Apple's been a huge champion of open standards. Everything from CUPs to USB and HTML5.

[smitty3268]

What's "as usual" supposed to mean exactly? Apple's been a huge champion of open standards. Everything from CUPs to USB and HTML5.

Apple is not known for having extremely spec compliant hardware.

They back whatever they want - sometimes it makes sense to use the standard, sometimes it doesn't. This time it didn't, and that's not exactly a surprise for anyone who follows apple.

[russofris]

Apple is not known for having extremely spec compliant hardware.

They back whatever they want - sometimes it makes sense to use the standard, sometimes it doesn't. This time it didn't, and that's not exactly a surprise for anyone who follows apple.

I think you might be misunderstanding the case here. Apple implemented a subset of the spec in firmware (cold plug) and the rest in software (hot plug). Is this really considered deviation? Aren't big bad firmware blobs what we should be against, and this is a case where Apple got it right?

[mjg59]

I think you might be misunderstanding the case here. Apple implemented a subset of the spec in firmware (cold plug) and the rest in software (hot plug). Is this really considered deviation? Aren't big bad firmware blobs what we should be against, and this is a case where Apple got it right?

Apple did this entirely correctly, and Intel's insistence on using SMM for every other implementation is unjustifiable at a technical level.

[russofris]

Apple did this entirely correctly, and Intel's insistence on using SMM for every other implementation is unjustifiable at a technical level.

I do not know whether Intel "insisted" that vendors use SMM, or was forced to provide SMM and leverage it because they perceived that MS/HP/Dell would resist kernel-side implementations in an effort to hamper Apple. Does anyone here have a grapevine to Jason Ziller, as he's probably the only one that could give us an informed answer?